Characterization problems in radio measurement systems

Sammanfattning: Radio measurement systems may have their performance significantly degraded due to environmental factors such as multipath, weather effects, and mechanical displacements. Characterization of these effects are therefore important in order to ensure functionality of the system. The characterization itself may also be the purpose of the system. This thesis contributes to the answer to the question of how to assess the effect of the environment on the propagation and reception of radio waves for three different applications.Traditionally the functionality of a radio measurement system has been assessed using either simulations assuming ideal conditions (e.g. free space) or measurements under controlled circumstances. There is no doubt that both these approaches are very useful when designing antennas and related hardware. In many applications it is also sufficient to assume ideal conditions and only use an a priori characterization. The applications considered in this thesis all operates in an environment that can be considered to be challenging. In these cases the environment needs to be taken into account in the design process of the system. Both simulations and measurement methods have been considered. The combination of electromagnetic simulation methods, such as the method of moments (MOM) or the partial element equivalent circuit (PEEC) method, with statistical methods, such as the Monte Carlo method, have been given special attention. The measurement systems considered, both for determining the performance of antennas and for detection of objects and transponders, are all assessed from a "challenging environment" point of view.The three application considered are multistatic radar using global navigation satellite systems (GNSS), measurement systems for antenna arrays in noisy conditions, and simulation of RFID systems with moving transponders. In the multistatic radar the focus is on detecting signals reflected in directions other than that of the primary reflection. The results shows that detecting these signals is possible even with the low signal levels involved. This is especially the case when reflecting objects are present which could scatter the signal in a specular way. By using the equivalent electric current method it is possible to estimate the complex far-field radiation pattern of antenna arrays even when the signals used have a low signal-to-noise ratio (SNR). This has been shown using simulation of a large antenna array and with measurements using a small array for GNSS receivers. When designing RFID systems it is important to be able to estimate the performance in terms of number of detected transponders with all movements of the transponders taken into account. This is possible by using a very simple model of the transponders (e.g. a magnetic dipole) in which case only one electromagnetic simulation is needed. This enables the use of the Monte Carlo method to take the random movements of the transponders into account using a low number of computations. The use of the PEEC method further enables a combined simulation of both the electromagnetic properties of the reader antenna and the electric functionality of the receiver circuit.Although the considered application are very different the obtained solutions are in many ways general. The fact that even the weak signals reflected in non-specular directions in a multi-static GNSS radar can be detected can be used in any application involving multi-path propagation or stray signals. The equivalent electric current method have here been considered for two radically different antenna arrays operating in a low SNR environment. Although the simulation approach chosen for the RFID simulations rely heavily on the simple magnetic dipole method it would work with any antenna at any frequency as long as the model of the antenna is sufficiently simple.